General Discussion

In 1983, owners’ equivalent rent replaced basic housing prices. The new formula was based on an estimate of what homeowners would have to pay to rent their own homes. But in real life, people pay mortgages, property taxes, and home prices. When home prices and mortgage rates rise faster than rents, the new formula understates the housing inflation real households face.

In 1996, the Boskin Commission announced that the Consumer Price Index (CPI) was overstated because people substitute lower-priced goods for higher-priced goods which are too slow in being calculated. The agency made the correction to eliminate the bias in the fixed basket of goods. The problem is that every single adjustment ended up forcing the reported rate to be less than a plain addition of the same goods over time.

In 1998, there was a new fashion for hedonic adjustments. This stemmed from an observation that quality is always improving, especially in digital goods and computer functioning. The idea is that you might be paying the same or even more but you are getting more bang for your buck with quality shifts. You guessed it: hedonic adjustments drew the inflation rate lower. Notably, hedonic adjustments never run the other way, raising prices when quality decreases. [Note: a repair technician once told me that refrigerators used to last 30 years but are often now kaput after 3.]

In 1999, a geometric mean formula replaced arithmetic mean for most CPI components. This was intended to capture substitution effects. This was the change that ended up disguising the increase in medical service costs. By looking at consumed services rather than actual prices, the inflation rate in this sector ended up burying inflationary trends. This highly technical adjustment completely ignored all the ways in which substitution is a behavioral adaptation to inflation, not a reduction in the inflation experienced.

In 2002, we got a continuation of this same method with new “chained CPI” which changes the basket weighting based on new purchasing patterns. Sure, if people buy less beef and more chicken, the household will experience inflation in a different way. But this ignores the manner in which the substitutions themselves are a response to higher prices. In 2017, the new calculation was applied to taxes causing people to pay more than they otherwise would have under the old method.

In 2018, the hedonic adjustment strategy was expanded to a huge new range of products including smartphones, residential telephone services, internet services, and cable and satellite television. In 2020, at the same time the composition of M1 was changed and not retrospectively applied such that the data is essentially useless. Following money supply data became more difficult. Then in 2024, the Bureau of Labor Statistics stopped looking at the actual cost of medical services and started only looking at claims, completing the consumption-only bias against actual posted prices [i.e., failing to capture the fact that many consumers had no insurance, etc]. In 2025, a month went by with no data collection at all.

So what happens when we strip all this away and examine actual prices as reported by the Bureau of Labor Statistics, without all the many adjustments? We find that a basket of goods and services that cost $100 in 1980 costs $515 per the Reality Index in 2025. The official CPI reports only $391.

That means that real prices have run 32% higher over 45 years than the government reports. Over a 55-year window, the Reality Index ran 54.4% faster than CPI.

To put it another way, consider the loss of purchasing power since 1980. According to the CPI, the loss has been to make $1 in 1980 worth only 26 cents. According to the Reality Index, the loss is greater: $1 in 1980 is now worth only 19 cents. By any standard, that is a shocking devaluation.